Chiral intermediates from aucubin as synthons of modified 11

Note: In lieu of an abstract, this is the article's first page. Click to increase image size Free ... The Journal of Organic Chemistry. Hagens, Miller...
0 downloads 0 Views 429KB Size
J. Org. Chem. 1982,47, 1343-1345 acetate/hexane. The 6337 isolated product ratio was matched by HPLC [ChromaneticsSpherisorb ODS reverse-phase column, methanol/acetonitrile/H20,0.01 M NH4S04,0.005 M (NH4)2C03 (21:1)] analysis of the crude reaction product. The alkaloids had physical constants corresponding to those of the products previously reported in our synthesis using the epoxy aldehyde 7.3 (b) From a reaction mixture sealed in glass and heated at 210 "C for 30 min after addition of triethylamine were isolated 42% of pandoline and 24% of 20-epipandoline (7030 product ratio by HPLC). (c) A reaction analogous to part a but using the bromo lactol 20a and with refluxing for 2 h after addition of triethylamine gave a 63:37 ratio of pandoline to 20-epipandoline. (d) A reaction at 20 "C for 24 h after addition of triethylamine gave equal amounts of pandoline and 20-epipandoline. (e) When 0.64 g (3.1 mmol) of the bromo lactol20a and 0.50 g (2.05 mmol) of the indoloazepine 6 in 20 mL of benzene were stirred under nitrogen at 47 "C for 72 h, most of the azepine reacted, and traces of pandoline and 20-epipandoline formed. Addition of 1 mL of triethylamine, continued stirring at 47 "C for 72 h, and a workup as under a gave 173mg (24%)of pandoline and 282 mg (39%)of 20-epipandoline. HPLC of the crude alkaloid mixture showed a 33:67 ratio. (f) An analogous reaction mixture stirred at 20 "C for 7 days after addition of triethylamine gave 176 mg (24%) of pandoline and 238 mg (33%)of 2O-epipandoline,with a 3664 ratio of isomers in the crude reaction product. (9) A solution of 950 mg (3.0 mmol) of the epoxy aldehyde 7 (at least 40% pure by NMR)3 and 400 mg (1.6 mmol) of the

1343

indoloazepine 6 in 20 mL of benzene was held at 47 "C under nitrogen for 4 days and then worked up as under a to give 110 mg (19%) of pandoline and 244 mg (42%) of 20-epipandoline. The crude reaction product showed a 31:69 HPLC ratio of these alkaloids.

Acknowledgment. Support for parts of this research project by the National Cancer Institute under National Institutes of Health Research Grant R01 CA 12010 is gratefully acknowledged. We are indebted t o Mr. W. Earley and Mr. T. Zebovitz of our group for respectively providing mass spectra and 250-MHz NMR spectra. Registry No. 1, 21217-98-1;2a, 73824-79-0;2b, 73805-38-6; 6, 66859-22-1; 7, 80641-93-6;8a, 80695-59-6; 8a.HC1, 80695-60-9; 8a acetate, 80695-61-0;8b, 80695-62-1;8b.HC1, 80695-63-2;8b 3,5-dinitrobenzoate, 80641-94-7; 8b acetate, 80695-64-3; 8b 14-thiocarbamate, 80641-95-8; 9, 74129-84-3; 10, 80779-26-6; 10 acetate, 80658-38-4; 1la, 80665-38-9; 14, 80642-07-5; 14 hydrazone, 8064208-6; 16, 80658-39-5;18, 18374-17-9;19, 80642-05-3;19 hydrazone, 80642-06-4;20,80641-96-9;22,80641-97-0;22a, 80641-98-1;22a hydrazone, 80641-99-2;ethyl 2-ethyl-4-pentenoate,80642-00-8;methyl 2-ethyl-4-pentenoate,42998-02-7;ethyl 4,5-epoxy-2-ethylpentanoate, 80642-01-9; methyl 4,5-epoxy-2-ethylpentanoate,74121-06-5; 2ethyl-4-pentenoicacid, 1575-73-1;2-ethyl-4-pentenoicacid p-toluamide derivative, 80642-02-0;5-chloro-2-ethyl-4-hydroxypentanoic acid 1,4-lactone,80642-03-1;methyl 4,5-dichloro-2-ethylpentanoate, 80642-04-2;14a-methyl-d-norvincadifformine, 80658-40-8;5-chloro4-ethyl-4-hydroxypentanoicacid 1,4-lactone,80642-09-7;&bromo4-ethyl-4-hydroxypentanoicacid 1,4-lactone,80642-10-0.

Notes Chiral Intermediates from Aucubin as Synthons of Modified 1 l-Methylprostaglandins. Assignment of Correct Structures to Two Tetrahydrodideoxyaucubins Carlo Bonini*' and Romano Di Fabio Centro di Studio per la Chimica delle Sostanze Organiche Naturali del CNR, Istituto di Chimica Organica dell'Universitci di Roma, 00185 Roma, Italy Received September 14, 1981

Current interest in the field of prostaglandins is focused on obtaining optically active intermediates without resolution. To this end many important syntheses starting from natural chiral compounds with appropiate asymmetric centers have been carried In particular, one industrial research group leaded by Ohno has recently patented a n original conversion of aucubin 1, the predominant naturally occuring iridoid glucoside, to 11deoxy-ll-a-(hydroxymethy1)prostaglandinF, and related compounds? Furthermore, Berkowitz e t aL5have carried out two syntheses of chiral prostaglandin intermediates (1) From Nov 1981 to Oct 1982 at the Department of Chemistry, Stanford University, Stanford, CA 94305. (2) G. Stork and S . Rauscher, J. Am. Chem. Soc., 98, 1583 (1976). (3) K. G. Paul, F. Johnson, and D. Favara,J.am. Chem. SOC., 98,1285

(1976). (4) K. Ohno,M. Naruto,N. Naruse, and H. Takeuchi, Tetrahedron Lett. 3,251 (1979);Jpn. Kokai Kokkyo Koho, No.83436 (1977); Chem. Abstr. 88, 6412 (1978). (5) W. F. Berkowitz, I. Sasson, P. S. Sampathkumar,J. Hrabie, S. Choudhry, and D. Pierce, Tetrahedron Lett. 1641 (1979);W. F. Berkowitz and S. Choudhry, Ibid., 1075 (1981).

from aucubin and asperuloside. With our recent interest in the iridoid chemistry! we have developed new routes' to prostaglandin synthons from aucubin, which is easily obtained from Aucuba japonica and Eucommia ulmoides. Our approach is outlined in Scheme I, where we describe our procedure t o obtain 8 and 9 which can be considered key intermediates in obtaining, by standard procedures, modified analogues of ll-methylprostaglandins. Our starting material was the known dideoxytetraacetylaucubin 3: obtained by Li/NHBreduction at -40 "C of aucubin 1 and subsequent acetylation of 2. To obtain compounds 4 and 5, we used a procedure described by Berkowitz5 which in the case of dideoxytetraacetylaucubin turned out t o be very efficient. The use of anhydrous MezSO yielded a bromohydrin-free mixture of bromo lactones 4 which were directly reduced (Zn/AcOH) without purification t o compound 5 (95% overall yield). The key step of the reduction of the double bond in compound 5 presented much difficulty, probably because of the bulky glucose moiety. We obtained good results by sterespecific reduction in the following manner: compound 6 (a-methyl group epimer at C,) was obtained by PtOz reduction in MeOH at (6) A. Bianco, C. Bonini, M. Guiso, C. Iavarone, and C. Trogolo, Tetrahedron Lett., 4829 (1978); Tetrahedron 73, 1773 (1981). (7) Another method not described in this paper has been recently patented by C. Bonini, R. Di Fabio, C. Iavarone, and C. Trogolo,Italian Patent Application 49042A/81 (1981). (8) A. J. Birch, J. Grimshaw, and H. R. Juneja, J. Chem. SOC.,5194 (1961).

0022-3263/82 / 1947-1343sO1.25,/O 0 1982 American Chemical Society I

,

1344 J. Org. Chem., Vol. 47,No. 7, 1982

Notes

Scheme I

-

1 R=OH.R'=H 2 R=R'=H 3 z R = H , R'=COCH, I I

4

R=Br

5

a=H

I

U

U

7

?.

/

8

?.

+

* \

Table I. Significant NMR Chemical Shift Values for Compounds 6, 7 , 10, and 11 compd carbons

CH,-10 c-9

6 15.72 43.16

7

10

11

18.74 47.94

18.51 43.10

19.54 49.64

4 atm, while the best conditions to obtain compound 7 (p-methyl group epimer) were use of P d / C in MeOH a t 1 atm.g T h e exact configuration of the methyl group at the C8 position in compounds 6 and 7 was demonstrated by 13C NMR comparison with the two knowna tetrahydrodideoxytetraacetylaucubinsto which we now may assign the configurations 10 (mp 141-142 "C) and 11 (mp 90-91 "C). See Table I for selgcted and significant spectral data. This was possible on the basis of specific studies of methyl substitution on the cyclopentane ringlo and several appropriate examples of 13CNMR data of compounds with related structures. In particular we cite two examples in recent literature,l1Pl2 in which the authors compared the 13C NMR chemical shifts of certain compounds having a methyl group in a cyclopentane ring cis or trans with respect to a cis ring junction. As given,11.12the chemical shift values of the a-methine carbon and of the methyl group in a cis relationship with a cis ring junction (CH, endo) (9)It is noteworthy that the catalytic effect we found without previous suggestionwaa reported by F. Murai and M.Tagawa [Chem.Phurm. Bull. 28(6),1730 (1980)]in the reduction of similar iridoid glucosides, which could represent a more general useful phenomenon. (10)M.Christl, H. J. Reich, and J. D. Roberta, J.Am. Chem. SOC.,93, 3463 (1971). (11)E. J. Eiaenbraun, C. E. Browne, R. L. Irvin-Willis, D. J. McGurk, E. L. Eliel, and D. L. Harris, J. Org.Chem., 45, 3811 (1980). (12)A. Bianco, P. Caciola, M. Guiso, C. Iavarone, and C. Trogolo, G a m . Chim. Ital., 111, 201 (1981).

are upfield with respect to the corresponding chemical shift values of the a-methine carbon and of the methyl group in a trans relationship with a cis ring junction (CH3 exo). Therefore, when considering compounds 6, 7, 10, and 11, the less deshielded CH3-10 and C-9 values can be assigned to the structures with the Me group in the CY configuration (CH3 endo): for compound 6, 15.72 and 43.16 ppm; for compound 10,18.51 and 43.10 ppm. On the other hand, the more deshielded CH,-10 and C-9 values can be attributed to the Me group in the @ configuration for compounds 7 (18.74 and 47.94 ppm) and 11 (19.54 and 49.64 ppm).13 The last step to compounds 8 and 9 was carried out in two different ways (MeO-/MeOH, K2C03/MeOH),both in good yields. Epimerization of the CHO group to the more stable @ configuration was completed, for compound 8, in a few minutes ['H NMR (CDC13) 6 9.56, (J = 4 Hz, d, CHO)]. For compound 9 the epimerization, which can be followed by observing the lH NMR resonances of aldehydic protons [6 9.56 (J = 4 Hz, d) and 9.75 (J = 2 Hz, d)] was almost complete after 20 h. It is noteworthy that, u p to now, 9,ll-deoxy-lla- or -1lp-methylprostaglandinsor appropriate synthons have been synthesized and that proper configurations have never been definitively assigned to compounds 10 and 11. Experimental Section Melting points are uncorrected. 'H NMR spectra were recorded on a Perkin-Elmer R-32 at 90 MHz, with Me4Si as an internal standard. 13CNMR spectra were recorded on a Varian CFT-20, with chemical shifts given in parts per million from Me4Sias an internal standard. IR spectra were run on a Perkin-Elmer Model 457.

Isolation of Aucubin. Aucubin was isolated from Aucuba japonica and Eucommia ulmoides according to the procedure described by Duff.14 The yield of crude aucubin from fresh plant was 2%. Preparation of Lactone 5. To a solutin of 3 (1.4 g, 2.9 "01) in dry and distilled (CaHJ M G O was added NBS (1 g, 5.8 "01) at room temperature with stirring under a nitrogen atmosphere. After 10 min the solution became yellow, to the reaction mixture was added water (50 mL), and the whole mixture was extracted with ether (200 mL). The etheral solution was then washed several times with water to remove the MezSO and then dried over Na2S04. Evaporation of ether gave 4 (1.6 g) as a solid residue which was used without purificatin in the next step. To a solution of 4 (1.6 g, 2.8 mmol) in ether (150 mL) were added with stirring acetic acid (8 mL) and then Zn (0.3 g, 5 "01). After 10 min the reaction mixture was diluted with ether (200 mL) and washed with a solution of saturated Na2C03and water. The etheral solution, dried over Na2S04and concentrated in vacuo, yielded a solid residue of 5: 1.35 g (95% overall yield from 3); mp 132-134 "C (EtOH); 'H NMR (CDCl,) d 5.52 (Hl, br s), 5.40 (H,, br s), 2.59 (H4,br d), 1.73 (CH,-10, br s). Anal. Calcd for C23H3001?: C, 55.42; H, 6.06. Found: C, 55.53; H, 6.01. Preparation of Lactone 6. A solution of lactone 5 (0.5 g, 1 mmol) in 50 mL of MeOH and 0.1 g of PtO, was hydrogenated at 25 O C and 4 atm of Hzfor 1 h. The reaction mixture was fiitered and evaporated to yield 6 0.45 g (90%);mp 144-146 "C (EtOH); lH NMR (CDC13)d 5.56 (Hl, d, J = 2 Hz), 2.06, 2.00, and 1.97 (OCOCHS, ~ , 1 H), 2 1.08 (CH3, d, J = 6 Hz); 13C NMR (CDC13) 99.02,97.06,72.87,72.35, 71.00, 68.35, 61.90,43.16,37.11,35.46, 33.77, 33.28, 32.90,20.60, 15.72 ppm. Anal. Calcd for C23H32012: C, 55.19; H, 6.44. Found C, 55.28; H, 6.37. Preparation of Lactone 7. A solution of lactone 5 (0.3 g, 0.6 mmol) in 40 mL of MeOH and a catalytic amount of 10% PD on activated charcoal was hydrogenated at 25 "C and atmospheric NMR data of CH,-10 epimeric iridoid gluco(13)See for further sides: (a) Y. Ozaki, S. Johne, and M. Hme, Helu. Chim. Acta, 62(8),2708 (1979);(b) F. Bailleul, P. Delaveau, A. Rabaron, M. Plat, and M. Kock, Phytochemutry, 16,723, (1977). (14)R. B. Duff, Biochem. J.,96,1 (1965).

J. Org. Chem. 1982,47, 1345-1347 pressure until absorption ceased (1 h). The reaction mixture, filtered and evaporated, gave a residue (0.3 g) which, purified by crystallization from EtOH, yielded 7 as a white solid mp 128-130 OC; lH N M R (CDCl,) 6 5.42 (Hl, br s), 2.1,2.02, and 2.0 (OCOCH3, 8, 12 H), 1.28 (CH3, d, J = 2 Hz; 13CNMR (CDClJ 100.38,97.25, 72.84, 72.35,70.99,68.26,61.89,47.94,38.27,34.76,34.36, 33.68, 33.21, 20.60, 18.74 ppm. Anal. Calcd for C23H32012: C, 55.19; H, 6.44. Found C, 55.25; H, 6.41. Methanolysis of Lactone 6 to 8. A solution of 6 (0.2 g, 0.4

mmol) in anhydrous MeOH was stirred at room temperature under a nitrogen atmosphere. To this solution was added dried After 10 min the reaction was completed, K2CO3 (0.2 g, 1.4 "01). and the solution was carefully neutralized with 0.1 N HCl. The methanol was then removed in vacuo at 35 OC, and the resulting aqueous solution was extracted several times with ether (100 mL). Evaporation of the ethereal solution extracts, dried over Na2S04, gave 8: 55 mg (75%);an oil; 'H NMR (CDC13)6 9.56 (CHO, d, J = 4 Hz), 3.63 (COOCH3,s), 1.05 (CH3,d, J = 6 Hz); IR (CC14) 2700, 1750, 1730 cm-'. Anal. Calcd for CloHleO3: C, 65.19; H, 8.75. Found C, 65.27; H, 8.69. Methanolysis of Lactone 7 to 9. A solution of lactone 6 (0.2 g, 0.4 mmol) in anhydrous MeOH was stirreed at room temperature under a nitrogen atmosphere. To this solution was added dried KzCO3 (0.2 g, 1.4 mmol), and the reaction was completed in 20 h. The reaction mixture was worked up as described above for 8, yielding 9: 50 mg (70%);an oil; 'H NMR (CDC13)6. 9.56 (CHO, d, J = 4 Hz), 3.63 (COOCH3, s), 1.26 (CH3, d, J = 2 Hz); IR (CC14)2700,1750,1735 cm-l. Anal. Calcd for C1OH16O3: C, 65.19; H, 8.75. Found: C, 65.23; H, 8.68. Preparation of Compounds 10 and 11. Compounds 10 and 11 were prepared accordingto the procedure described by Birch.8 For compound 1 0 mp 140-142 "C (lit. mp 141-142 "C); 13CN M R (CDCl3)94.75,94.49,72.76,71.58,71.00,68.28,61.68,58.57,43.10, 33.19, 32.61,32.29, 29.89, 27.05, 20.18, 18.51 ppm. For compound 11: mp 89-90 OC (lit. mp 90-91 "C) '% NMR (CDC13)95.08,94.88, 73.13,71.90, 71.32,68.69,62.06,58.89,49.68,33.52,32.94,31.19, 30.22, 27.37, 20.51, 19.54 ppm.

1345

addition of trifluoroacetic anhydride or trifluoroacetyl chloride to the substrate or by the in situ generation of the acid anhydride from trifluoroacetic acid (TFA) and a dehydrating agent, e.g., tetraphosphorus decaoxide? The formation of trifluoroacetic acid esters would be anticipated from the reaction of the acid with an alcohol or phenol. The practical need for a nonreversible reaction dictates for many acids the use of the acid chloride or the acid anhydride; however, a number of esters have been isolated in this study without the use of these reagents. Quantitative yields of the ethyl, n-butyl, benzyl, methyl, and 2-naphthyl esters are obtained by refluxing the hydroxyl compound with commercial TFA. A notable exception is that phenol requires the addition of tetraphosphorus decaoxide to the reaction mixture. The phenol ester is the most rapidly hydrolyzed of the esters isolated in this study. The cleavage of benzyl and of tert-butyl ethers by TFA has been noted;68 however, the general use of this reagent for cleaving ether linkages has not been demonstrated. Several ethers have been found to cleave in this study. The most probable reaction route would have an oxonium ion intermediate since TFA is a very strong acid; K, = 0.5 in acetic acidg (eq 1). The complex then decomposes with CF3C02H

+ n-C4HioOC2H5

-

[n-C4H,oO(H)C2H,I [02CCF,I (1) cleavage of the ether linkage. The decomposition products can form by a substitution process or by elimination (eq 2 and 3). The initially formed substances then undergo

[~-C~H~OO(H)C~H~I[~~CCF~] -+

n-C$IioOH

+ C2H502CCF3 (2)

[~-C~H~~O(H)C~HSI[~~CCFBI -.+

Acknowledgment. We thank S. L. Frachey and F. Piccioni for accurate measurement of 'H NMR and 13C NMR spectra. Registry No. 3, 19467-17-5; 4, 80359-98-4; 5, 80359-99-5; 6, 80360-00-5;7,80408-21-5;8,80360-01-6;9,80408-22-6;10,80408-23-7; 11, 80408-24-8.

Reaction of Trifluoroacetic Acid with Alcohols, Phenols, Ethers, and Their Sulfur Analogues Arnulf P. Hagen,* Tammy S. Miller, Richard L. Bynum, and Ved P. Kapila Department of Chemistry, The University of Oklahoma, Norman, Oklahoma 73019 Received July 3, 1981

n-C,HloOH

Results and Discussion The importance of trifluoroacetate derivatives of biochemical and organic substances has been well establ i ~ h e d . ~ The ? ~ derivatives are normally formed by the (1) Miller, T. S. M.Sc. Thesis, The University of Oklahoma, 1980. (2) Hagen, A. P.; Bynum, R. L.; Kapila, V. P.; Miller, T. S. Fuel, in

press. (3) Heuser, E. "Cellulose Chemistry"; Wiley: New York, 1944; pp 230-232.

0022-3263/82/1947-1345$01.25/0

(3)

their characteristic reactions including polymerization, addition, and esterification. In this specific reaction the major products are the ethyl and n-butyl esters of TFA. Ethyl and diphenyl ethers do not cleave in a sealed tube at temperatures up to 180 "C; however, n-butyl ethyl ether cleaves a t 72 "C to form the ethyl and n-butyl esters of TFA. Bis(1-phenylethyl) ether cleaves to initially form an alcohol and an ester (eq 4). The alcohol then splits out [C6H&H(CH&],O + CF3C02H CGH~CH(CHJOH CGH,CH(CH3)OC(O)CF, (4) -+

+

a molecule of water to form styrene which polymerizes. The water hydrolyses the ester to give TFA and additional alcohol which loses water. Therefore the overall reaction becomes that of eq 5.

TFA

When bituminous coal is treated with trifluoroacetic acid (TFA) at 72 "C it forms a powder having greatly reduced sulfur and ash contents. This report describes the reaction chemistry of TFA with pure substances having functional groupings that would be anticipated for bituminous coal.'f

+ C2H4 + CF3C02H

[CGH&H(CH~)I~O

(CGHSCHCHJ,+ H2O (5)

Benzyl ether reacts to form poly(pheny1ene methylene) analogous to the reaction of benzyl alcohol with concentrated sulfuric acid.1° The TFA reaction can take place via an oxonium complex which decomposes to form benzyl alcohol and an ester which then undergo further reactions. (4) Haschemeyer, R.; Haschemeyer, A. 'Proteins: A Guide to a Study by Physical and Chemical Methods"; Wiley: New York, 1973; p 362. (5) Sleevi, P.; Glass, T. E.; Dorn, H. C. Anal. Chem. 1979,51, 1931. (6) Beyerman, H. C.; Heiszwolf, G. J. Red. Trau. Chim. Pays-Bas1965, 84, 203. (7) Beyerman, H. C.; Bontekoe, J. S. Red. Trau. Chim.Pays-Bas1962, 81, 691. (8) Marsh, J. P.; Goodman, L. J. Org. Chem. 1965, 30, 2491. (9) Howells, R. D.; McCown, J. D. Chem. Rev. 1977, 77, 69. (10) Shriner, R. L.; Berger, A. J. Org. Chem. 1941, 6, 305.

0 1982 American Chemical Society